JP3565391B2 - Viewpoint correction autostereoscopic display device - Google Patents

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to an autostereoscopic display device.
[0002]
[Prior art]
In normal operation of looking at an object, the two human eyes perceive an external view in their respective perspectives. This is due to the distance between eyes (interocular distance). The brain then uses these two views to estimate the distance of various objects in a scene. In order to obtain a display device that effectively displays three-dimensional (3D) images, it is necessary to reproduce this situation and supply so-called “stereo pairs” of images one by one to each eye of the observer.
[0003]
Depending on how different views are supplied to each eye, 3D displays are categorized into two types: stereoscopic and autostereoscopic. Typically, stereoscopic displays display both images over a large viewing area. In one known type of configuration, the observer wears two head mounted separate channels, such as head mounted displays. Each channel gives each of the autostereoscopic pairs of images an associated eye. In other types of stereoscopic displays, both images are typically displayed over a wide viewing area. FIG. 1 shows such a display device 1 and a wide output light cone 2 obtained by the display device 1. Each view is coded for color (in the case of analog systems), polarization state, or time (in the case of shutter glasses systems). The observer wears visual aids such as filters 3 and 4 in front of the right and left eyes R and L so that the views are separated so that only the intended view of each eye is visible. As shown in FIG. 1, the left and right views are coded into codes A and B, respectively. The filter 3 such as a color filter, a polarizing filter, or a shutter blocks the light having the code A and transmits the light having the code B, so that the right eye R can see the right-eye view. Similarly, the filter 4 blocks the light having the code B and transmits the light having the code A, so that only the left eye view is seen by the left eye.
[0004]
In the case of an autostereoscopic display device, the observer does not need to wear any visual aid. Instead, as shown in FIG. 2 of the accompanying drawings, the two views are only visible from a limited spatial area. The autostereoscopic display device 1 creates “visual areas” such as 6 and 7. The visual region is a spatial region in which a single two-dimensional (2D) image is visible to one eye over the entire active area of the display device 1. When the observer is in a situation where the right eye R is in the right visual area 7 and the left eye L is in the left visual area 6, an autostereoscopic pair of images is visible and a 3D image is perceived.
[0005]
In the case of a flat panel autostereoscopic display device, a visual region is typically formed by cooperation of a picture element (pixel) structure of the display device and an optical element called a parallax optic. The parallax optical system includes a parallax barrier, a lenticular screen, a hologram, and the like. Parallax barriers are screens with translucent vertical slits separated by opaque areas. FIG. 3 shows a front parallax barrier type autostereoscopic display device. The parallax barrier 8 is disposed in front of a glass substrate 10 and a spatial light modulator (SLM) 9 including a column of pixels 11 (hereinafter, columns). There is a gap 12 between columns of adjacent pixels 11. The SLM 9 may be a light-emitting device such as a pixelated electroluminescence display device, but as shown in FIG. 3, the SLM 9 is a light valve type such as a liquid crystal device (LCD) and has a backlight 13. Have.
[0006]
The pitch of the slits 14 is selected to be close to an integral multiple of the pitch of the columns of the pixels 11. Thereby, the pixel column group is associated with each slit of the barrier 8. As shown in FIG. 3, the slits 14 are associated with three columns 1, 2, and 3, respectively.
[0007]
The function of the parallax optical system (parallax barrier) 8 is to restrict the direction in which light passes through each pixel to a predetermined output angle range. For primary light, the angular range of view for each pixel is determined by the width of the pixel and the spacing between the pixel plane and the parallax optics 8 plane.
[0008]
FIG. 4 of the accompanying drawings shows the angular bands of light created by the SLM 9 and the parallax barrier 8. Here, the pitch of the parallax barrier 8 is exactly an integral multiple of the pitch of the column of the pixels 11. Angle bands Z1 and Z2 of light emitted from different positions on the display device surface mix with each other. Thus, there is no area in front of the display where a single image is visible to one of the observers throughout the display surface.
[0009]
In order to overcome this problem, as shown in FIG. 5, the pitch of the parallax optical system 8 is slightly reduced so that the angle zones Z1 and Z2 are focused on a predetermined plane in front of the display device. This plane is referred to as a “window” plane 15. The change in the pitch of the parallax optical system 8 is referred to as “viewpoint correction”. From “viewpoint correction”, visual regions 6 and 7 are obtained. Within the visual regions 6 and 7, Z2 and Z1 of all angle zones respectively overlap (overlap) each other. In general, the visual areas 6 and 7 extend in the vertical direction and have the shape of a “kite” in a horizontal plane.
[0010]
The window plane 15 defines the best viewing distance of the display device. The best performance of the display is obtained when the observer's eyes are in the window plane 15. As each of the observer's eyes moves laterally in the window plane 15, there is no change in the perceived image until the eyes reach the sides of the viewing area 6 or 7. As the eye moves to an adjacent visual area, the image perceived across the display device changes (eg, to the next image). Each visual area 6 and 7 portion of the window plane 15 is commonly referred to as a "visual window".
[0011]
In the case of a typical SLM, such as a thin film transistor liquid crystal display (TFT LCD), the columns of pixels 11 are separated by gaps 12 to obtain electrical connection paths. The gap 12, which forms a vertical strip, is covered by an opaque material to block light leaking from the gap. In the case of a TFT LCD, this opaque layer is called a "black mask" or "black matrix". However, as shown in FIG. 6 of the accompanying drawings, the vertical strips between the columns of pixels 11 also image in the window plane, thereby forming a low brightness area 16 between the visual areas 6 and 7. In order to prevent the formation of the low-luminance area 16, that is, to make the visual areas 6 and 7 contact each other, an adjacent set of pixel columns related to each parallax element of the parallax optical system (parallax barrier) 8 is set horizontally The openings in the black mask that define the pixels 11 must be formed so that they are seamless in direction (ie, there is no continuous vertical black mask strip between adjacent sets of pixel columns).
[0012]
The illumination profile (variation in light intensity depending on the visual position) in each of the visual regions 6 and 7 is determined by the shape of the opening defining the pixel 11. Since the parallax optical system (parallax barrier) 8 is a cylindrical optical element, the vertical openings of each pixel column are integrated to provide illumination extending vertically in each of the visual regions 6 and 7. Thus, as shown at 17 in FIG. 7, when the width of the vertical aperture of the pixel varies, the intensity of the illumination changes along the width of the visual window. This is shown in FIG. The visual region 7 is divided into a high-luminance band 18, a low-luminance band (dull zone) 19, and a mixed band 20. When the eyes move between the high luminance band 18 and the low luminance band 19, the observer perceives an intensity change on the order of 5% or more as a visual flicker effect. This effect is offensive to degrade the perceived display quality. Therefore, in such a display device, it is desirable to keep the ratio of the vertical openings constant by using a rectangular pixel opening or the like.
[0013]
If the observer's eyes are not located in the window plane 15 and there is a defect in viewpoint correction, different information will be visible at different locations on the display device surface. For example, if the observer's eyes are in the mixing zone 20 near the display, the observer will see the left side of the display substantially brighter than the right side of the display. If the observer is far enough away from the window plane to move out of the viewing areas 6 and 7, different eye slices of each eye will be visible on the display surface. For this reason, the 3D effect is lost. This situation begins to occur at the ends of the visual areas 6 and 7 which are closest to the display and which are farthest from the display. Low brightness bands due to vertical strips between pixels appear on the display as lower brightness bands.
[0014]
As shown by columns 1, 2, and 3 in FIG. 3, each parallax element is primarily associated with each pixel column group, and adjacent pixel column groups are also imaged by this element. As shown in FIG. 8 for a two-view display device that displays views V1 and V2, the columns are imaged to form a lobe consisting of a visual region repeated on both sides of a central (ie, zero-order) lobe. Is created. Each of these lobes repeats all of the characteristics of the central lobe, but is more susceptible to optical system defects and aberrations. Therefore, higher order lobes may not be available.
[0015]
Generally, to obtain a full color display, each pixel 11 is optically matched to a filter associated with one of the three primary colors (red, green, blue). By properly controlling the group of three pixels associated with the three primary color filters, nearly all of the visible colors can be generated or simulated. In order to obtain a balanced color output using an autostereoscopic display, each stereoscopic image channel must contain sufficient color filters. To simplify manufacturing, many SLMs have color filters arranged in vertical columns. Thus, every pixel in each column has the same color filter associated with it. When a parallax optical system is arranged in such an SLM and three pixel columns are associated with each parallax element, light that forms an image in each visual region is monochromatic. Therefore, the configuration of the color filter must avoid this situation (EP0752610).
[0016]
The autostereoscopic display device shown in FIGS. 3 to 7 has a parallax barrier as the parallax optical system 8 in front of the display device (that is, between the SLM 9 and the visual regions 6 and 7). However, the parallax optical system having another configuration has substantially the same operation. For example, as shown in FIG. 9, instead of the front parallax barrier, a front lenticular screen having an array of lens pieces (that is, lenticules) that converge light in a cylindrical shape may be used. The lenticular screen focuses light from the SLM 9 on the window plane and provides a visual area with a well-defined border area on axis. Lenticules do not limit the throughput of light, as in the case of parallax barriers, but instead redirect light, so lenticular screens have higher illumination at the window plane. However, there is no problem with the parallax barrier, but there is a problem with the influence of optical aberration caused by the lenticular screen.
[0017]
In the autostereoscopic display device shown in FIG. 10 of the accompanying drawings, a parallax barrier 8 is arranged between the backlight 13 and the SLM 9 to form a rear parallax barrier display device. Is different from that shown in This configuration has an advantage that the parallax barrier 8 is hardly damaged. This is because the parallax barrier 8 is held behind the SLM 9. If the rear surface of the parallax barrier 8 is made to reflect light so that light not incident on the slit can be reused (instead of being absorbed), light efficiency can be improved. The switchable diffuser 21 is disposed between the SLM 9 and the parallax barrier 8, and may include a polymer-dispersed liquid crystal or the like. When this display device is switched to the low diffusion state, it operates as an autostereoscopic 3D display device as described above. When the diffuser 21 is switched to a highly diffused state, light rays are deflected as they pass through the diffuser, forming an even (or "Lambertian") distribution. An even distribution prevents the generation of visual bands. Therefore, the display device functions as a 2D display device, and can display a 2D image using all the spatial resolutions of the SLM 9.
[0018]
FIG. 11 shows a known type of spatial light modulator (SLM) 9. The SLM 9 is a liquid crystal display (LCD) including a plurality of picture elements (pixels) arranged in a regular pattern or array in a matrix. The LCD (SLM) 9 includes a red pixel 32, a blue pixel 33, and a green pixel 34, and performs color display. The LCD 9 is a thin film transistor twisted nematic type, and pixels are separated from each other by a black mask 35. Thus, each column of pixels is separated from the adjacent column by a vertically continuous, opaque black mask 35 strip. The black mask 35 prevents light from passing through the thin film transistors of the LCD (SLM) 9.
[0019]
In order to perform 3D display, a lenticular screen 8 is arranged in front of the pixels of the LCD 9. The lenticular screen 8 has a plurality of lenticules extending in the vertical direction, and each lenticule converges light in a cylindrical shape. The lenticules extend vertically and may be formed, for example, as plano-convex cylindrical lenses or graded index (GRIN) cylindrical lenses. Each lenticule is located above a plurality of pixel columns (four columns in FIG. 11), each column of pixels providing a vertically sliced 2D view piece. The shape of each pixel is a shape obtained by combining a rectangle with a small rectangular extension protruding from the right side of each pixel.
[0020]
As shown in FIG. 12, the 3D display device is appropriately illuminated from the rear so that each column of pixels displays a 2D image piece sliced vertically thin. In this case, if the image data is provided to the pixels of the LCD 9, each lenticule of the lenticular screen 8 forms a visual band 37 to 40 from the four pixel columns associated with the lenticule, respectively. The direction in which the visual bands 37 to 40 spread corresponds to the direction in which each 2D view was recorded during the period in which the image was captured. A 3D image is perceived when each of the observer's eyes is within adjacent visual bands 37-40.
[0021]
FIG. 13 shows a 3D display of the type disclosed in EP 0 625 861 with an LCD 9 and a lenticular screen 8. The LCD 9 differs from the LCD shown in FIG. 11 in that pixels are arranged in a pattern consisting of horizontal rows (hereinafter, rows) and vertical columns. In particular, each pixel may be a composite pixel having a red pixel 32, a blue pixel 33, and a green pixel 34. The pixels are arranged without any break in the horizontal direction. In other words, there is no vertically continuous black mask section separating the pixels. To achieve this, each composite pixel 50 in the first row is vertically spaced from a composite pixel 51 in the second row adjacent in the horizontal direction. However, the right side of the composite pixel 50 is on the same vertical line (the direction in which the pixel columns extend) as the left side of the composite pixel 51. Therefore, as compared with FIG. 11, the vertical resolution of the LCD 9 is practically halved, but the number of pixel columns imaged by each lenticule of the screen 8 is doubled to eight.
[0022]
As shown in FIG. 14, each lenticule of the screen 8 generates eight visual bands 52-59. These output light beams are angularly continuous and represent eight different 2D views with continuous horizontal parallax. Thus, there is no “black” region like 41 and “grey” region like 46 shown in FIG. 12, and the observer perceives a 3D image with substantially constant intensity and no image gap. be able to. Furthermore, the number of 2D views for each 3D image frame is doubled by halving the vertical resolution.
[0023]
EP0617549 discloses a head-mounted stereoscopic display device having a separate display device and an optical system for each observer's eye. Each display device includes a backlight and an LCD, and each pedocal system forms a virtual image for a left-eye view or a right-eye view of the stereo pair. For comfortable vision, a virtual image is formed in the same region in front of the observer.
[0024]
EP 0 262 955 discloses an autostereoscopic display of the type providing two views repeated in a plurality of lobes.
[0025]
[Problems to be solved by the invention]
However, in the apparatus shown in FIGS. 11 and 12, the vertical portion of the black mask 35 between the pixel columns also forms an image in the directions indicated by 41 to 45. Further, the visual bands 37 to 40 include regions with reduced brightness, such as 46 to 48. The decrease in brightness occurs because a rectangular projection extending from the main pixel portion forms an image. For this reason, there is a problem that the output of the display device does not have continuous parallax of uniform brightness.
[0026]
The LCD 9 shown in FIG. 13 overcomes the shortcomings of the LCD 9 shown in FIG. 11 in that it can provide a continuous visual band. However, the pixels must be accurately and seamlessly connected in the horizontal direction to prevent unwanted visual artifacts from being seen by an observer. In particular, if there is an underlap (gap) or overlap (overlap) between pixels in the horizontal direction, when the observer's eyes move from each visual zone to an adjacent visual zone, intensity variations occur. Therefore, in order to avoid such an effect, this type of LCD must be manufactured with very strict tolerances, and there is a problem that the manufacturing becomes complicated and the manufacturing cost increases.
[0027]
In addition, as described below, in the LCD of FIG. 13, crosstalk between left and right views can cause undesirable visual artifacts. Specifically, there is a problem that the amount of visible crosstalk is different from each other or changes stepwise according to the movement of the observer.
[0028]
SUMMARY OF THE INVENTION The present invention has been made to solve the above problems, and an object of the present invention is to provide an autostereoscopic display device with less visual artifacts.
[0029]
[Means for Solving the Problems]
The viewpoint correction autostereoscopic display device of the present invention, At least one display device, and optics that cooperate with the display device to form at least three viewing windows in a window plane such that adjacent pairs overlap laterally; An observer tracking system for determining an observer position and an image controller responsive to the observer tracking system, wherein the image controller comprises: Said Splits the image displayed by the display device ,Observer The window containing the left eye receives the view data for the left eye, the window containing the right eye of the observer receives the view data for the right eye, and the adjacent window exists when the observer's eye is in the area where the adjacent windows overlap. One receives black view data Is to control Thereby, the above object is achieved.
[0045]
Preferably, the observer tracking system is configured to determine the position of a plurality of observers, and at least three windows are provided for each observer.
[0046]
More preferably, when the observer's eyes are in an area where the adjacent windows overlap, the image data received by one of the adjacent windows is changed from black. Corresponding to the observer's eyes Switch to image data, and at the same time, Corresponding to the observer's eyes The image controller is configured to switch from image data to black.
[0047]
In addition, the viewpoint-correcting autostereoscopic display device of the present invention includes at least one display device and, in cooperation with the display device, at least three visual windows on a window plane such that adjacent pairs overlap laterally. An optical system formed in An observer tracking system for determining the position of the observer, and an image controller responsive to the observer tracking system, wherein the image controller comprises: display Equipment display In the area on the surface Images from each of the above Supplied to the window Split In an area where the observer's left eye perceives only left-eye image information, the observer's right eye perceives only right-eye image information, and the observer's eyes receive light in two of the windows, One of the two windows is switched to black So that the above object is achieved. .
[0048]
Preferably, Said The display device and the optical system Working together, Assuming that e is the average interocular distance and N is the number of windows for each lobe, a window having a lateral pitch substantially equal to (2 × e / N) is formed.
[0049]
The picture element may have a shape such as a rectangle or a parallelogram. According to such a configuration, it is possible to avoid variations in illumination intensity on the visual window.
[0050]
The spatial light modulator may be embodied as a light-emitting device such as an electroluminescent display device or a light-transmitting device such as a liquid crystal device in association with an illumination source such as a backlight.
[0051]
The parallax device includes a parallax barrier, a lenticular screen, a hologram, and the like.
[0052]
The picture element may have a shape such as a rectangle or a parallelogram. According to such a configuration, it is possible to avoid variations in illumination intensity on the visual window.
[0053]
The spatial light modulator may be embodied as a light-emitting device such as an electroluminescent display device or a light-transmitting device such as a liquid crystal device in association with an illumination source such as a backlight.
[0054]
BEST MODE FOR CARRYING OUT THE INVENTION
ADVANTAGE OF THE INVENTION According to this invention, the display apparatus which does not require a movable part at all and can track an observer in a horizontal direction or in the horizontal direction and the distance direction from an apparatus can be provided. A lenticular screen or a front or rear parallax barrier may be used for this display. With three visual windows per lobe, the loss of resolution in 3D mode of the display can be minimized. Alternatively, three or more windows can be used to enhance certain aspects of display performance. By providing at least three windows for each observer, several observers can be tracked separately, if desired. In pixelated devices, pixel boundaries do not introduce undesirable visual artifacts, thus allowing for manufacturing tolerances. Further, the crosstalk performance of the display device is improved as compared with a known display device. For example, the amount of each visible crosstalk is substantially the same, and does not change stepwise according to the observer's movement.
[0055]
If a liquid crystal device is used, the device can be manufactured using existing technology with only minor modifications. For example, a suitable device can be obtained simply by modifying the black mask of a known type of LCD.
[0056]
The present invention will be further described with reference to embodiments with reference to the accompanying drawings. In the drawings, the same members are given the same reference numerals.
[0057]
FIG. 15 shows a part of the LCD 9. The LCD 9 differs from the LCD shown in FIG. 13 in that pixels 50 overlap in the horizontal direction. The pixel 50 is a rectangle whose sides are aligned in the row direction and the column direction. The width w of each pixel is greater than the pixel pitch p in the horizontal direction (ie, the row direction), which results in an overlap, represented by m, between each adjacent pair of pixels.
[0058]
The parallax device 8 is shown as a lenticular screen, but may be any other suitable device such as a parallax barrier. The pitch P of the parallax device 8 is substantially equal to an integer multiple of the pixel pitch p. However, as described above, in the case of a view point corrected display, the pitch P of the lenticular screen 8 is made slightly smaller than an integral multiple of the pitch p of the pixels 50.
[0059]
The 3D display device shown in FIGS. 13 and 15 with back lighting can be used in a reversionary 2D mode. In the 2D mode, when the LCD of FIG. 15 is used for such a display device, the display device becomes brighter than the display device using the LCD of FIG. However, if the vertical range of pixels and the width of the vertical black mask are the same, the brightness in the autostereoscopic mode is the same.
[0060]
A display device of the type shown in FIG. 15 can be used in a projection display system. For example, the output of the display can be projected behind a parallax barrier or lenticular screen. Alternatively, when several projectors are imaged on a direction maintaining screen such as an autocollimation reflector or a retroreflector, overlapping windows are obtained by overlapping the openings of the projector lens.
[0061]
The display device of FIG. 15 may be a color display device. For example, a color filter arrangement of the type disclosed in EP0625861 or EP0752610 can be used for the LCD of FIG.
[0062]
FIG. 16 shows the visual band created by imaging pixel 50 through one of the lenticules of screen 8. When the LCD 9 is properly illuminated by a backlight or the like, the lenticule produces a fan-shaped set of nine visual bands 53-61 as shown. Adjacent visual band pairs, such as 53 and 54, have an overlap at an angle corresponding to the overlap region m of adjacent pixels. The LCDs shown in FIGS. 15 and 16 generate three visual zones in each of the three lobes as shown. Pixels 50 located just below the lenticules create visual bands 56-58 in the zeroth lobe. Through this same lenticule, the three pixels below the adjacent lenticule create visual bands 53-55 in the + 1st lobe and the three pixels 50 under the adjacent lenticule on the opposite side. Produces visual bands 59-61 in the -1st order lobe.
[0063]
FIG. 17 illustrates visual artifacts that may be caused by an LCD of the type shown in FIG. Due to manufacturing tolerances, adjacent pixels 50 may not be connected strictly in the horizontal direction. FIG. 17 shows a case where there is an underlap (i.e., a gap between adjacent pixels) caused by such a tolerance. As indicated by the optical path from the observer's eye 62, the lenticular screen 8 has an on-axis resolved spot size. This on-axis resolution spot size is actually several micrometers. As a result of manufacturing tolerances that cause an error of the order of one micrometer on pixel continuity, as the eye image ("eye spot") in the LCD pixel plane crosses the boundary between adjacent pixels, this manufacturing tolerance Appears as a change in light intensity from the lenticule. This is shown in FIG. 17 for the case of underlap, and the intensity profile of the light intensity corresponding to the position of the eye is indicated by 63. The intensity decreases as the eye 62 crosses the boundary between adjacent pixels. On the other hand, when there is an overlap between pixels, the intensity increases. Improving manufacturing tolerances can reduce such strength variations, but it increases manufacturing costs and makes manufacturing more difficult.
[0064]
As shown in FIG. 18, the lenticular screen 8 forms an eye spot as a bar 64 having a fine width. The light intensity is proportional to the area of the overlap between the bar 64 and the pixel 50. Since the underlap exists as shown in FIG. 18, the overlap area changes when the bar 64 crosses the boundary between adjacent pixels.
[0065]
By providing an overlap area m between adjacent pixels 50 and switching between adjacent pixels when the observer's eyes are within the overlap area of the adjacent visual band, the LCD 9 in FIG. Overcoming the problem. If adjacent pixels have the same intensity performance, ensure that the overlap area m is wider than the eye spot, and switch between pixels when the eye spot is completely within the overlap area. , Visual artifacts due to changes or variations in intensity can be substantially eliminated or made invisible. Small intensity changes due to intensity mismatches between adjacent pixels can be substantially invisible by crossfading between adjacent pixels, thus reducing or substantially eliminating visual artifacts. . In addition, manufacturing tolerance requirements are substantially reduced accordingly. This is because there is almost no need to confirm whether the width of the overlap region m provided between adjacent pixels is sufficient.
[0066]
In manufacturing a black mask for a liquid crystal display, it is difficult to obtain very sharp corners in the holes. Thus, at the corners of a substantially rectangular pixel, there is an associated curve of a certain radius. This causes a loss of vertical holes and a slight reduction in brightness on the sides of the window. In the case of the display device of FIG. 15, if the viewer does not cross the side of the window during a normal tracking operation, all possible visual artifacts caused by this decrease in brightness can be avoided.
[0067]
FIG. 19 shows an autostereoscopic display of the type disclosed in EP 0 726 482 with the lenticular display shown in FIG. When the image data is provided to the display device, a number of visual bands are generated, each corresponding to a different view. The visual zones corresponding to the same view converge to a predetermined position, thereby forming a viewpoint correction zone. In the viewpoint correction zone, the observer can observe the autostereoscopic effect. The widest part in each viewpoint correction zone defines a “window”. A "window" appears at a predetermined distance from the display device.
[0068]
The windows are seamlessly connected to each other and define a laterally extending viewing area. If the image displayed in each window is updated according to the position of the observer, an autostereoscopic effect is observed in this visual region. The three windows labeled 1, 2, and 3 image into three lobes labeled -1, 0, and +1. In the case of the illustrated "three-window" display device, the horizontal extent of each window is two-thirds of the average interocular spacing of the observer group. The display device is configured so that each window displays either left view data or right view data. Left and right views are not mixed in one window.
[0069]
When the observer is at the position A in FIG. 19, the observer's right eye R is in the first window 1 of the 0th order lobe, and the observer's left eye L is in the second window 2 and the third window 3 of the 0th order lobe. On the border of. The positions A to G indicate the lateral position of the observer, but the position in the distance direction of the observer is always at a normal visual distance. Control the first window 1 (via the first display device) to display the right view data, and control the second and third windows (via the second and third display devices) to display the left view data. Display and get an autostereoscopic view. As the observer moves from position A to position B, the observer's right eye moves to the boundaries of the first and second windows of the zeroth lobe. Similarly, the observer's left eye also moves away from the boundaries of the second and third windows of the zero order lobe. As a result, since the second window 2 is no longer observed, the image data displayed in the second window 2 is updated from the left view data to the right view data in anticipation that the observer will reach the position B. be able to. When the observer comes to the position B, the position of the observer's right eye is on the boundary between the first window 1 and the second window 2 of the 0th-order lobe. The first and second windows both display right view data. On the other hand, the position of the left eye of the observer is located at the center of the third window 3 displaying the left view data.
[0070]
When the observer moves from the position B to the position C, the observer predicts that the left eye of the observer will be able to observe the first window 1 of the + 1st lobe, and updates the first window 1 of the + 1st lobe. To display the left view data.
[0071]
FIG. 20 schematically illustrates an autostereoscopic display device including the LCD and the lenticular screen shown in FIG. In the figure, the display device 15 generates three visual windows (denoted as 1 to 3) by repeating two lobes. Various positions in the lateral direction of the observer are indicated by AG. A to G represent horizontal positions on the window plane. The table shows image data displayed in three windows at each position of the observer. The observer tracking system determines the position of the observer and controls the image data supplied to the three image display devices. In the configuration of the lenticular display device illustrated in FIG. 15, the pixels are denoted by 1 to 3 in order to indicate which image display device the pixel configures. In this type of display, the images are spatially multiplexed as interleaved image strips formed by pixel columns. There are three pixel columns located below each lenticule on the screen 8.
[0072]
When the observer is at the position A, the position of the left eye is within the window 3, so that the left view data is displayed by the pixels forming the third image display device. The right eye is in the overlap area between window 1 and window 2. When the observer moves to the left, the right-eye view data displayed in window 1 is switched to window 2, and the pixels supplying the image to window 1 are switched to black. Therefore, at position B, the right and left view data are displayed in windows 2 and 3, respectively.
[0073]
When the observer passes position C, the right eye remains in window 2. Therefore, window 2 continues to display the right image view data. However, the left eye moves to the overlapping area of window 3 and window 1 of the adjacent lobe. Therefore, the left image view data is supplied to the pixels forming the window 1, and the pixels forming the window 3 are switched to black so that no image is displayed. Thus, as previously described, visual artifacts caused by variations in intensity as the observer moves as compared to the continuous-type display shown in FIG. 13 are substantially reduced or eliminated. Is done.
[0074]
The lateral tracking operation of the display device shown in FIG. 13 is described in EP 0726482. This operation differs from the operation described with reference to FIG. 20 in that at any timing, two adjacent windows display one view of the autostereoscopic pair, and the third window displays the other view of the autostereoscopic pair. The view of is displayed. FIG. 21 shows the crosstalk performance due to this operation. The upper part of FIG. 21 shows the relationship between the position and the intensity when the window 1 and the window 2 display the left image data. The middle figure shows the window 3 displaying the right image data. As shown in the lower diagram of FIG. 21, light from a window leaks into adjacent windows on both sides. However, since two windows display the left view and one window displays the right view, the amount of light leaking from windows 1 and 2 to window 3 as compared to the amount of light leaking from window 3 to windows 1 and 2 The amount is large. Therefore, the right view contains more crosstalk than the left view.
[0075]
The situation is reversed when the observer moves, and the crosstalk occurring in the left view is greater. As a result, some image flicker artifacts are visible when the observer moves.
[0076]
FIGS. 22 and 23 show the effect of crosstalk between the left and right views in the display device shown in FIG. FIG. 22 shows the crosstalk immediately before the image data is switched, and FIG. 23 shows the crosstalk immediately after the switching. In the case of this display device, since only one window is illuminated at any time for each view, the amount of crosstalk in both the left and right views is substantially equal. Therefore, image flicker artifacts due to crosstalk variations are substantially eliminated.
[0077]
The best overlap of adjacent windows, that is, the best overlap of adjacent pixels, is a compromise between the crosstalk performance and the aberration performance of the display. FIG. 24 shows an eye spot obtained as a vertical bar by a lenticular screen or parallax barrier. Eye spots L1 and R1 indicate the position of the eye spot when the observer is at the first position, and eye spots L2 and R2 correspond to the observer at the second different position. As shown, the width of the eye spot is larger than the width of the overlap area between adjacent pixels. In this case, as schematically shown in the graph that is a part of FIG. 24, the observer sees a change in intensity as the eye passes over the overlap region. Also, the eye spot increases as the observer moves away from the axis of the display device. Thus, the flicker artifact increases as the observer moves away from the center position, thereby limiting the allowable visual freedom for the display device. On the other hand, the spacing between “active” windows is further increased, thus reducing crosstalk between views.
[0078]
FIG. 25 shows a case where the overlap between adjacent pixels 50 is relatively large. Since the width of each eye spot is substantially smaller than the width of the overlap region between adjacent pixels, intensity variations are substantially reduced or eliminated. However, the distance between the active windows displaying the left and right image view data is greatly reduced. Thus, crosstalk increases because light leakage between the active windows is substantially increased. When the width of the eye spot is smaller than the gap between adjacent pixels, visible crosstalk is minimal. When the eye spot becomes large due to the effect of aberration at a position distant from the axis of the display device, the amount of visible crosstalk increases. This can cause variations in intensity as the observer moves, which can result in other flicker-type artifacts. Therefore, the selection of the width of the overlap region between adjacent pixels is made to obtain an acceptable compromise between intensity variation and crosstalk variation.
[0079]
Providing a display of the type shown in FIG. 15 having four columns of pixels 50 under each lenticule of the lenticular screen 8 results in a display that produces four visual windows per lobe. . Such an arrangement is much less susceptible to the above-described error mechanism (immune to the error mechanism). This is because at any position, both eye spots need not be above the same view pixel. To achieve this, the visual windows are arranged in such a position that switching of the view data occurs almost simultaneously for the left and right image view data. In order to prevent the variation in intensity shown in FIG. 24, the overlap area between adjacent pixels may be increased, but the variation in crosstalk performance may occur when the observer moves in the lateral direction with respect to the display device. The duplicate window is always switched to black so that there is no window.
[0080]
FIG. 26 schematically illustrates a video multiplexing system that controls the views displayed by the image display device. Three or more windows are created, but only the left-eye and right-eye view information is needed. The left-eye view input of the first, second, and third video switches 102, 104, and 106 is provided with left-eye view information via the buffer 100. The right eye view input of the first, second, and third video switches is provided with right eye view information via buffer 108. Black view information is provided to the black inputs of the first, second, and third video switches via the buffer 109. Each video switch is responsible for selecting a video view to be provided to one of the three image display devices for display in one of the windows. Each video switch may control its own display device, or each video switch may drive a single display device in a multiplexed manner.
[0081]
Each video switch receives a control input from the controller 110. The controller 110 selects which of the left view data, the right view data, and the black view data is to be displayed. The controller 110 is responsive to a tracking system 112 that determines the position of the observer. The controller 110 obtains information on the observer's position and display parameters, selects an appropriate view based thereon, and commands the video switch to display the appropriate left or right view, or And Alternatively, controller 110 may respond to manual control 111. The manual control 111 is manually controlled by the observer to perform manual observer tracking.
[0082]
FIG. 27 schematically illustrates an autostereoscopic display device that performs observer tracking in a lateral direction and a distance direction using the LCD 9 and the lenticular screen 8 shown in FIG. The optical path from the display to the observer is shown. The observer is at a position closer to the display device than the predetermined position (that is, between the display device and the window). By following this optical path from the window border, through each of the observer's eyes and towards the display, the image seen by the observer can be calculated. For the right eye, the image of window 3 is in area 200 of the display, the images of windows 2 and 3 are in area 201 of the display, the image of window 2 is in area 202, and the images of windows 1 and 2 are It can be seen in the area 203. To the left eye, the images of windows 1 and 2 are visible in area 204, the images of window 1 are in area 205, and the images of windows 1 and 2 are in area 206 of the display device.
[0083]
FIG. 28 shows a possible selection example of selecting the image content of the window so as to maintain the view of the autostereoscopic image. The right-eye information is displayed in windows 3 and 2 in regions 207 and 208 of the display device. The left-eye information is displayed in window 1 of area 209, window 2 of area 210, and window 3 of area 211 on the display device. In this manner, the information displayed by the pixels forming the windows 1 to 3 is sliced so that only the left-eye information can be seen by the left eye and only the right-eye information can be seen by the right eye. Such an image slicing method is disclosed in EP0721131.
[0084]
If the observer is farther away from the display and the visual window is located between the display and the observer, a similar display analysis and display control is used to ensure autostereoscopic vision. can do. The visual region in which the observer can perceive the 3D image can be enlarged in the lateral direction and the distance direction.
[0085]
When there are three windows for each lobe, the display device is configured such that the average interocular distance of the observer is approximately equal to 1.5 times the window pitch at the best visual distance. (Actually, when the number of windows for each lobe is N, the general formula is Wp = 2e / N. Wp is the pitch of the windows and e is the average interocular distance of the user group.) It corresponds to the distance between the eye spots on the pixel plane. The interval between the eye spots on the pixel plane is substantially equal to 1.5 times the pixel pitch p. The distance between the eye spots increases as the observer approaches the display device, and decreases as the observer moves away from the display device. The maximum distance between eye spots that enables autostereoscopic vision is 2p, and the minimum distance is p. The maximum and minimum visual distance of a lenticular display device can be calculated as follows.
[0086]
In a display device having three windows, a normally focused lenticular screen having a thickness of t, a refractive index of n, and an interocular distance between observers is e , The normal visual distance Z corresponding to the window position _nom Is expressed as follows.
[0087]
Z _nom = Et / (1.5np)
The maximum and minimum visual distances are expressed as:
[0088]
Z _min = Et / (2np)
Z _max = Et / (np)
For a display device with N windows, the normal, maximum, and minimum viewing distances are expressed as:
[0089]
Z _nom = (2 ^* e ^* t) / (n ^* N ^* p)
Z _nim = Z _nom (1- (N-2) / (2 (N-1))
Z _max = Z _nom (1+ (N-2) / 2)
As described above, if the number of visual windows is increased, visual flexibility in the distance direction is improved. If more windows are used, the degree of visual freedom in the horizontal direction in an actual display device is also improved. This is because the eye spot switching point is located farther from the pixel boundary, and the effect of aberration on image quality is small at the switching point.
[0090]
FIG. 29 schematically illustrates a video multiplexing system for generating sliced video image fragments. Although three or more windows are provided, all that is required is left-view and right-view view information. The left eye view information of the first, second, and third video switches 302, 304, and 306 is provided to the left eye view input via the buffer 300. Right-eye view inputs of the first, second, and third video switches are provided with right-eye view information via a buffer 308. Black view information is provided to the black inputs of the first, second, and third video switches via a buffer 309. Each video switch is responsible for generating a video view that is provided to one of the image displays to generate a view in one of the windows. Each video switch may control a respective display device, or, as shown in FIG. 27, each video switch may drive a single display device in a multiplex manner.
[0091]
Each video switch receives two control inputs from controller 310. The controller 310 selects which of the left view data, the right view data, or the black view data is to be displayed in a portion of the video output. The controller 310 is responsive to a tracking system 312 that determines the position of the observer. The controller 310 obtains information of the observer's position and display parameters, selects an appropriate view based on the information, and gives an instruction to the video switch to display the corresponding portion in the left view, the right view, or black. . For example, the view display device shown in FIG. 29 corresponds to image data necessary for the operation shown in FIG.
[0092]
The display device can be configured to track more than one observer and provide each observer with a 3D image within the expanded visual range. For example, the display device shown schematically in FIG. 30 can track two observers. The LCD has seven sets of pixels arranged as overlapping pixel columns under each lenticule of the lenticular screen. Thus, this display provides seven windows with overlap. The seven windows can be repeated in adjacent lobes. This display device can simultaneously track two observers to perceive a 3D image.
[0093]
The display device shown in FIG. 31 is a beam combiner type using the small illuminator 13, and is different from the embodiments described above. Each of the small illuminators 13 includes a lenticular screen. A parallax barrier is arranged behind the lenticular screen. The parallax barrier has a plurality of slits. The slits are each aligned with one of the lenticules of the lenticular screen. An even diffusion backlight is located behind the parallax barrier.
[0094]
Each lenticule of the lenticular screen images a parallax barrier slit aligned with the lenticule into a window of the zero order lobe. The slit also forms an image of the −1st and + 1st lobes of the visual window in the same window by the adjacent lenticules. As described above, the display device shown in FIG. 31 performs observer tracking by the same operation as described above.
[0095]
The display device, a part of which is shown in FIG. 32, uses a parallax barrier 8 instead of the lenticular screen, and is different from that shown in FIG. The barrier 8 has a plurality of slits. These slits are aligned with the pixels of LCD 9 as shown in FIG. A Lambertian backlight 13 is provided behind the LCD. LCD pixels have overlap. Although the pixels are shown on different planes for clarity, they are typically arranged on a single plane.
[0096]
As shown in FIG. 32, each slit directs light along the light cone from the associated pixel to the first, second, and third windows of the zero order lobe. In addition, these slits create + 1st and -1st order lobe windows such that the windows of each lobe overlap each other and the lobes overlap each other.
[0097]
A display of the type shown in FIG. 32 with a “front” parallax barrier 8 produces a darker image at certain illumination levels than a display using a lenticular screen. However, the parallax barrier is not affected by optical aberrations unlike a lenticular screen. In the case of the front parallax barrier display device, the quality of windows generated in the observer plane (particularly, the boundary width of each window) is controlled by the slit width. Reducing the width of the slit reduces the geometric spread of each window side. However, when the width of each slit is reduced, the amount of diffraction at the slit increases. Therefore, it is necessary to select the slit width as a compromise between the diffraction effect and the geometric degrading effects.
[0098]
The display device partially shown in FIG. 33 uses a rear parallax barrier 8 instead of the front parallax barrier 8, and is different from that shown in FIG. The rear parallax barrier 8 is arranged between the LCD 9 and the backlight 13. A switchable diffuser 21 is provided between the barrier 8 and the display device 9. The surface of the barrier 8 facing the backlight 13 is made to reflect light. Thereby, light that does not pass through the slit of the barrier 8 is reflected and returns to the backlight 13 to be reused. Therefore, the brightness of the display image is improved.
[0099]
If a rear parallax barrier 8 is used, the geometric extent of each window side is controlled by the slit width of the parallax barrier, while the diffractive spreading of the window side is determined by the pixel width of the SLM 9. Controlled. Therefore, as compared with the display device shown in FIG. 32 using the front parallax barrier, the quality of image formation on the window can be improved. The effect of diffraction is disclosed in British Patent Application No. 96254977.4.
[0100]
By switching the switchable diffuser 21 to diffuse the light from the slits in the barrier 8, the display device can be used for 2D operation. In this case, if the LCD 9 is illuminated with a Lambertian source, a 2D image can be viewed over a wide visual range.
[0101]
A rear parallax barrier 8 may be formed by providing a transparent slit array in an opaque mask. Alternatively, the barrier may be formed by imaging a defined size light source on the diffuser through a lenticular screen.
[0102]
FIG. 34 schematically illustrates another technique for generating three or more windows with overlap by hologram 131. The hologram 131 has a plurality of holographic elements 132. A plurality of holographic elements 132 are each associated with a pixel of the spatial light modulator of the display and are tuned for an appropriate color filter of the pixel. Such holographic elements 132 are equivalent in operation to a lenticular screen or parallax barrier, and when properly illuminated by a parallel white light reproducing beam 133 or the like, each holographic element 132 is defined. ) Create a window of the relevant color. Each holographic element can be recorded to define several lobes as shown in FIG. Holographic elements 132 are arranged in groups, whereby light from each pixel group is imaged into one of the three or more window groups shown in FIG. The light intensity is controlled by the pixel whose attribute and directionality are switched by the hologram 131. One of the benefits of using holograms is that the display performance at off-axis locations is greatly improved. This is because the aberration at a position away from the axis is almost canceled when recording the hologram.
[0103]
As shown in FIG. 36, a hologram 131 may be arranged inside the SLM 9 together with the liquid crystal layer 136 and the color filter 137. Therefore, the hologram is substantially arranged on the plane of the liquid crystal device forming the SLM 9 by a method such as controlling a black mask pattern in the pixel hole. The hologram of each pixel can be adjusted to direct light of a particular color associated with that pixel's color filter to the appropriate window. This is shown in abbreviated form in FIG. "W" means window, "R", "G", and "B" mean red, green, and blue light, respectively. Therefore, the white light performance of the display device can be improved. For example, a light source for a display may include three narrow spectral peaks. By appropriately selecting the phosphors used in the fluorescent tube such that the spectral spread of the light from the holographic element is relatively small when combined with color filters and pixels, a spectral peak is obtained.
[0104]
FIG. 37 shows another configuration in which the hologram 131 is located on the outer surface of the SLM 9. The use of collimated illumination in this configuration makes use of holograms formed in photopolymers or dichroic gelatin, or holograms in the form of etched surface reliefs or embossed holograms. Is possible. Alternatively, it is possible to create a diffraction grating in the SLM by controlling the electrode configuration in each pixel of the SLM 9.
[0105]
Hologram 131 can be generated by a computer or recorded by interfering light from an illuminated window with a reference beam. For example, using a mask and a red reference beam, it is possible to expose the recording plate through a red filter at the location of the first window of each lobe. Next, this process is repeated for green light and blue light. Thereafter, this process is repeated for each window and each corresponding holographic element.
[0106]
FIG. 38 shows a parallel white small backlight that can be used to illuminate the hologram 131. This backlight has a barrier 57 arranged between the elements of the lens array 54. The lens array 54 is arranged so as to obtain parallel light. The rear surface of the barrier 50 is made to reflect light. As a result, the unused light is returned to the backlight illuminator 13 and reused.
[0107]
The parallel backlight shown in FIG. 39 is different from that shown in FIG. 38 in that a small glass sphere 140 having a large packing density is used instead of the lens array 54 and the barrier 57. In yet another embodiment, it is possible to obtain parallel light by an edge light hologram.
[0108]
In this way, an observer tracking autostereoscopic display device having no moving parts can be obtained. Such a display device is tougher and quicker to respond than a display device using a movable part. In addition, such a display device has a relatively large tolerance for an error during tracking by an observer (relatively insensitive).
[0109]
Non-mechanical / electronic, lateral / distance tracking function and mechanical tracking method (moving parallax element such as parallax barrier or lenticular screen directly to SLM, or at least parallax element and It is possible to combine the functions of a flat panel display device with an SLM, such as rotating a sandwich (sandwich). Therefore, in addition to the speed and expanded visual freedom obtained by the non-mechanical tracking method, good aberration performance is obtained by the mechanical tracking method (image of the observer's eye in the SLM plane through the parallax barrier) Is maintained at or near the center of the SLM pixel). In this way, as the observer moves to another position, the relatively slow mechanical system causes the parallax barrier or lenticular screen to move directly relative to the SLM or to rotate the stack. be able to. Further, when the non-mechanical tracking method is linked with the mechanical tracking method, the observer can maintain the autostereoscopic image with the visual freedom expanded in the distance direction. This is not possible using only mechanical methods.
[0110]
An LCD 9 of the type shown in FIG. 15 may be manufactured by a special manufacturing method so as to obtain a required pixel pattern having an overlap, but such manufacturing is relatively uneconomical. Specifically, as new mask patterns for LCD active elements and internal electrodes are required, new electronic drive schemes are required.
[0111]
FIG. 40 shows a conventional "delta pattern" black mask of a known type. This black mask defines the active pixel area that is visible in a conventional RGB panel display. This mask can be modified to define overlapping apertures for active pixels, thereby creating the LCD of FIG. Except for this point, the LCD is conventional, but as described above, the LCD is suitable for use in an autostereoscopic display. In particular, there is no need to change the arrangement of addressing electrodes such as transistors and diodes or active matrix elements in the display device. Also, there is no need to move transistors and other electronic components that form part of the electronic drive scheme of the display. Therefore, it is not necessary to provide a completely new mask set for an active matrix display device.
[0112]
If the color filters of the conventional delta pattern display device are omitted, an LCD suitable for use in a monochromatic autostereoscopic display device can be obtained. In that case, three independent visual windows are obtained using the original red, green, and blue channels. Thus, three spatial multiplexed views can be displayed by providing the original red, green, and blue channels of the LCD with the pixels of the spatial multiplexed view. Thus, there is little or no need to modify the electronic drive system of the LCD.
[0113]
Thus, an LCD of the type shown in FIG. 15 can be manufactured using existing processes with little modification. Therefore, such an LCD is economical.
[0114]
FIG. 41 shows a display device including the spatial light modulator 40, the lens 77, and the illumination source 13. Illumination source 13 comprises six light emitting elements arranged in pairs represented by 72, 73, and 74 with overlap. By the action of the lens 77, an image of the light source 13 is formed at the normal visual distance Z. The distance between the elements of each light emitting element pair is the same as that of the other pairs, and these elements are arranged on a common plane. These light emitting element pairs are sequentially illuminated. As each illuminator pair illuminates sequentially, video information is sequentially provided to the spatial light modulator in a time multiplexed manner. With such a display, two overlapping lobes are obtained at the normal viewing position Z. Each lobe has three overlapping windows. If the six light emitting elements are individually controllable, it is possible to operate this display as a six window single lobe display to achieve an equivalent degree of freedom of movement.
[0115]
In the display device shown in FIG. 42, the first light source 13a includes three illuminators arranged at regular intervals, and is configured to illuminate the first spatial light modulator 9a via the lens 82. Have been. Similarly, the second light source 13b includes three illuminators arranged at regular intervals, and is configured to illuminate the second spatial light modulator 9b via the lens 86. The configuration is the same for the third light source 13c, the third lens 89, and the third spatial light modulator 9c. After being modulated by the respective spatial light modulators 9a and 9b, the respective images of the first and second light sources 13a and 13b are combined by a first beam combiner 90. The combined image is further combined with the image of the third light source 13c by the second beam combiner 92 after being modulated by the third spatial light modulator 9c. These images are arranged laterally offset from one another, so that the output of three overlapping lobes is obtained at the normal viewing position Z. The three lobes have three overlapping windows.
[0116]
FIG. 43 schematically shows a display device according to an embodiment of the present invention. Spatial light modulator 9 is sandwiched between first and second lenticular arrays 182 and 184. The first array 182 is near the spatial light modulator 9 and its pitch is approximately equal to the pitch of the spatial light modulator. The pitch of the second lenticular array 184 is approximately twice the pitch of the first lenticular array. The diffuser 186 is located between the spatial light modulator 9 and the second lenticular screen 184. First and second light sources 13 a and 13 b with overlap are arranged to illuminate first lenticular array 182 via lens 192. The diffuser 186 is arranged so that the images of the light sources 13a and 13b are formed on the diffuser 186 after the modulation by the spatial light modulator 9. The diffuser 186 is also in the object plane of the second lenticular screen 184. The second lenticular screen 184 re-images the diffuser 186 at the normal viewing position Z.
[0117]
The light sources 13a and 13b and the spatial light modulator 9 are driven by a time multiplex method. When the first light source 13a is illuminated, the first and second modulation elements 194 and 196 of the spatial light modulator 9 form an overlapped modulated image in the first region of the diffuser 186. When the first illuminator 13a is turned off and the second illuminator 13b is illuminated, the modulators 194 and 196 have an overlap in a second area on the diffuser 186 that overlaps the first area. Form an image. These images are re-imaged to form overlapping windows. In such an embodiment, both a spatial multiplexing and a time multiplexing are combined to obtain a multi-lobe four-view display device.
[0118]
FIG. 44 shows a projection display device including a plurality of projectors (only two are shown). Each projector includes a light source / reflector 400, a condenser lens 401, an LCD 402, and a projection lens 403. The image displayed by the LCD is projected onto an autocollimating screen 404, resulting in a window set 405 with overlap. Window 405 with overlap is an image of the projection lens aperture in each lobe. The lobes are obtained by the autocollimation screen 404.
[0119]
FIG. 45 shows an arrangement of projection holes that can be used to form a window 405 having an overlap. Because these holes are circular, there is a variation in intensity on the overlapping windows. As shown in FIG. 46, if the projection holes are limited using a square mask, the intensity in the window plane can be made more uniform.
[0120]
Various modifications can be made without departing from the scope of the invention. For example, instead of using a SLM to properly illuminate the display device, other types of image display devices, such as luminescent or reflective display devices, can be used.
[0121]
The SLM 9 shown in FIG. 15 has a regular array of rectangular pixels 50, which have constant vertical holes. However, other configurations are possible. For example, FIG. 47 shows an array of parallelogram pixels where the pixels have a constant vertical hole.
[0122]
In a display device that performs electronic tracking, it is important that the intensity level of the visual window be uniform so that the observer can move without perceiving undesirable visual artifacts. In this way, the observer can move within a certain visual area or move to another visual area displaying the same image without seeing a ticker or a change in display intensity.
[0123]
One problem that can occur with the SLM 9 described so far, as shown at 500 in FIG. 48, is that when the SLM displays a low brightness or black background, the pixel boundaries appear as thin white lines. There is that. Such light leakage is considered to be due to the polarization action. In an LCD switched to black, light is blocked by crossing polarizers and cannot be transmitted. This input light is polarized in one plane by the input polarizer. This plane is re-oriented by the liquid crystal to be 90- to the preferred transmission axis of the output polarizer. Thus, substantially all light is blocked by the second polarizer, and the display looks black.
[0124]
If light leaks from the side of the pixel in the black state, there must be a disturbance of the polarization plane near the side of the panel. It has been found that when a pixel side is oriented at 45 ° with respect to the input (and output) polarizer, most of the light leaks from the pixel side. If the sides are oriented in the axial direction of either the input or output polarizer, no light leakage will occur. Unfortunately, in a typical TFT twisted nematic LCD, the polarizer is oriented at 45 ° with respect to the vertical and horizontal directions for viewing angle reasons. For this reason, a vertical side of a rectangular pixel or the like causes light leakage. This mechanism is considered to be due to the rotation of the polarization angle due to the reflection, diffraction, or scattering of light on the side of the pixel hole defined by the black mask 35.
[0125]
In the case of a 3D autostereoscopic display with a lenticular screen, light from the right edges 500 of the pixels is collected by the lens aperture and imaged into the window plane, thereby producing a thin vertical strip of light. . These strips are relatively visible against the low brightness image background and appear as thin light strips.
[0126]
FIG. 49 shows an example of a technique for reducing or eliminating such light leakage in the SLM 9. The SLM 9 includes glass substrates 10a and 10b. An input polarizer 503 is arranged between the substrate 10b and the backlight 13. A black mask layer 35 is formed on the inner surface of the substrate 10a. An active layer 502 having a liquid crystal layer, electrodes, and the like is provided between the substrates 10a and 10b. An output polarizer, typically located on the outer surface of substrate 10a, is internally located at 501 between black mask 35 and active layer 502. According to this configuration, since the light from each pixel is extinguished before reaching the black mask 35, the influence of the subsequent rotation of the polarization direction is eliminated, and no light leakage occurs.
[0127]
50 and 51 illustrate another technique for reducing or eliminating light leakage. The input polarizer 503 and the output polarizer 501 are oriented in respective polarization directions parallel to the sides of the pixel hole. The corner between adjacent sides of the pixel hole may be rounded as shown in FIG. 50, and in this case, light leakage can occur only at the rounded corner. However, when a second black mask 35b with parallel opaque strips is overlaid on the first black mask 35a, this combination forms a black mask 35 with sharp corners. As described above, rounding of the corner of the pixel due to a defect at the time of manufacturing can be almost eliminated, and light leakage at the corner can be substantially avoided.
[0128]
When using twisted nematic liquid crystals, aligning polarizers 501 and 502 in this manner affects the viewing angle of the LCD. As a result, as is well known, the contrast of the panel differs on both sides of the center line. This can be corrected by adding a birefringent (ie, reacted mesogen) layer 504, as shown in FIG. 51, or by using another liquid crystal material.
[0129]
This rotation of the polarization is due to the tilted light rays reflected on the sides of the black mask, and the use of a parallel light backlight can reduce this light leakage source. However, this affects the visual band of the display device and is difficult in small systems.
[0130]
When the light leakage mechanism is based on reflection, the amount of reflected light can be reduced by using a black mask having a relatively low reflectance. This effect can be reduced by using an organic pigment layer or an emulsion layer instead of a known inorganic layer.
[0131]
【The invention's effect】
According to the present invention, it is possible to provide a margin for manufacturing tolerance. This is because small errors in the position of the visual window do not result in undesirable visual artifacts.
[0132]
According to the present invention, observer tracking can be performed without providing a movable unit. The simplest form of such an observer tracking display device provides three windows on one display device.
[0133]
According to the present invention, the visual freedom of the observer is further increased. Preferably, the lobes overlap each other.
[0134]
According to the present invention, a pixelated spatial light modulator is available and embodies a display device in a sophisticated manner.
[0135]
According to the invention, the picture elements may also be in the shape of a rectangle or a parallelogram. According to such a configuration, it is possible to avoid variations in illumination intensity on the visual window.
[Brief description of the drawings]
FIG. 1 is a schematic plan view of a known type of stereoscopic display device.
FIG. 2 is a schematic plan view of a known type of autostereoscopic display device.
FIG. 3 is a schematic transverse sectional view of a known type of autostereoscopic display device.
FIG. 4 is a schematic plan view showing a non-viewpoint correction autostereoscopic display device.
FIG. 5 is a schematic plan view showing a known viewpoint correction autostereoscopic display device.
FIG. 6 is a view similar to FIG. 5, showing the effect of gaps between pixel columns in a display device of the type shown in FIG.
FIG. 7 is a view similar to FIG. 5, but showing the action of pixels of non-constant vertical holes in a display device of the type shown in FIG.
FIG. 8 is a schematic plan view showing lobe formation in a known type of autostereoscopic display device.
FIG. 9 is a schematic plan view showing another known type of autostereoscopic display device.
FIG. 10 is a schematic transverse sectional view showing another known type of autostereoscopic display device.
FIG. 11 is a view showing a known type of directional display device.
FIG. 12 is a diagram illustrating an output of the display device of FIG. 11;
FIG. 13 is a diagram showing another known type of directional display device.
FIG. 14 is a diagram showing light output of the display device of FIG.
FIG. 15 is a diagram showing a directional display device according to an embodiment of the present invention.
16 is a diagram showing the light output of the display device of FIG.
FIG. 17 is a diagram illustrating a variation in intensity caused by the display device of FIG. 13;
18 is a diagram showing a variation in intensity caused by the display device of FIG.
FIG. 19 is a diagram showing a known type of autostereoscopic display device that performs observer tracking.
FIG. 20 is a diagram illustrating an autostereoscopic display device according to an embodiment of the present invention that performs lateral observer tracking.
21 is a diagram illustrating a crosstalk effect that occurs in a display device of the type illustrated in FIG.
FIG. 22 is a diagram illustrating crosstalk performance of the display device of FIG. 20;
FIG. 23 is a diagram illustrating a crosstalk performance of the display device of FIG. 20;
FIG. 24 is a diagram showing the effect of changing the overlap of adjacent pixels on intensity fluctuations and crosstalk in the display device of FIG. 20;
FIG. 25 is a diagram showing the effect of changing the overlap of adjacent pixels on the fluctuation of intensity and crosstalk in the display device of FIG. 20;
26 is a diagram showing a control system for controlling the display device shown in FIG.
FIG. 27 is a diagram illustrating an autostereoscopic display device and an operation when an observer moves from a predetermined visual distance.
28 is a diagram illustrating a case where the display device illustrated in FIG. 27 is operated to perform observer tracking in a distance direction.
FIG. 29 is a diagram showing a control system of the display device of FIG. 28.
FIG. 30 shows a display device of the type shown in FIG. 15 for performing multiple observer tracking.
FIG. 31 is a diagram showing another autostereoscopic display device having overlapping windows.
FIG. 32 is a schematic sectional view showing a part of a display device using a front parallax barrier.
FIG. 33 is a schematic sectional view showing a part of a display device using a rear parallax barrier.
FIG. 34 schematically illustrates the formation of a visual window using a hologram.
FIG. 35 schematically illustrates the formation of a visual window using a hologram.
FIG. 36 is a schematic sectional view showing a part of a display device using an internal hologram.
FIG. 37 is a schematic sectional view showing a part of a display device using an external hologram.
FIG. 38 is a schematic sectional view showing a small backlight for illuminating a hologram.
FIG. 39 is a schematic cross-sectional view showing a small backlight for illuminating a hologram.
FIG. 40 shows a known type of delta pattern liquid crystal display device.
FIG. 41 is a schematic view showing a time multiplex display device according to an embodiment of the present invention.
FIG. 42 is a schematic view showing a three-view beam combiner display according to an embodiment of the present invention.
FIG. 43 is a schematic view showing a four view display device according to an embodiment of the present invention.
FIG. 44 is a schematic view showing a projection display device according to an embodiment of the present invention.
FIG. 45 is a view showing the arrangement of projection holes of the display device of FIG. 44.
FIG. 46 is a view showing the arrangement of projection holes of the display device of FIG. 44.
FIG. 47 is a diagram showing a pixel configuration using a parallelogram.
FIG. 48 is a view showing a state of the LCD SLM in a black state, and shows the occurrence of light leakage.
FIG. 49 is a cross-sectional view of an SLM configured to reduce light leakage.
FIG. 50 is a diagram showing a black mask arrangement for reducing light leakage.
FIG. 51 is a sectional view of an SLM using the black mask arrangement of FIG. 50;
[Explanation of symbols]
9. LCD with overlapping pixels
8 Parallax barrier
13 Small backlight
10a, 10b Substrate glass
21 Switchable diffuser
35 Black mask layer
35a first black mask layer
35b Second black mask layer
54 Lens Array
57 Barrier
102 Video Switch
110 controller
112 Tracking System
111 Manual control
132 Holographic element associated with each pixel and tuned for appropriate color filters
133 parallel light, white light
134 window with overlap generated in several lobes
136 liquid crystal layer
137 Color filter
131 Hologram
140 (Small diameter sphere with high packing density)
404 Auto Collimation Screen
Light leakage at the side of 500 pixels
501 Internal Polarizer (Output Polarizer)
503 input polarizer
504 Birefringent layer to improve viewing angle

Claims (5)

At least one display device;
Optics cooperating with the display device to form at least three visual windows in the window plane such that adjacent pairs overlap laterally;
An observer tracking system that determines the position of the observer,
An image controller responsive to the observer tracking system,
The image controller, wherein the dividing the image displayed by the display device, a window containing the left eye of the observer receives left eye view data, a window receives the right-eye view data including the right eye of the observer, the adjacent window Is a viewpoint correction autostereoscopic display device, which controls so that one of the adjacent windows receives the black view data when the observer's eyes are in an area where. The viewpoint-correcting autostereoscopic display device according to claim 1 , wherein the observer tracking system is configured to determine a plurality of observer positions, and wherein at least three windows are provided for each observer. When the observer's eyes are in an area where the adjacent windows overlap, the image data received by one of the adjacent windows is switched from black to image data corresponding to the observer's eyes ,
The viewpoint-correcting autostereoscopic display device according to claim 1 , wherein the image controller is configured to simultaneously switch image data received by the other of the adjacent windows from image data corresponding to the eyes of the observer to black. . At least one display device;
Optics cooperating with the display device to form at least three visual windows in the window plane such that adjacent pairs overlap laterally;
An observer tracking system that determines the position of the observer,
An image controller responsive to the observer tracking system,
The image controller divides an image in an area on a display surface of the display device so as to be supplied to each of the windows, a left eye of the observer perceives only left-eye image information, and a right eye of the observer. Is a viewpoint-corrected autostereoscopic display that controls only one of the two windows to be switched to black in an area where the observer perceives only right-eye image information and the observer's eyes receive light in two of the windows. apparatus. The display device and the optical system cooperate to form a window having a lateral pitch approximately equal to (2 × e / N) , where e is the average interocular distance and N is the number of windows per lobe; The viewpoint correction autostereoscopic display device according to claim 1 .

Images